Strangers were holding the buy box on their own product.

Where they started

Losing the sale on their own listing

Thunderbird’s problem was not demand. It was that when a customer arrived at the listing and pressed buy, the order frequently went to somebody else. Losing the buy box does not look like a crisis on a dashboard — sessions hold up, the listing is still there, the brand still appears — and that is exactly why it runs for months before anyone names it. The old report puts the cost at more than $50,000 in revenue.

The advertising had the same quality of problem. The initial audit found thousands of dollars a month going to placements that were not returning, on an account nobody had rebuilt in a while. Money was moving. Nothing was compounding.

What we did

Take the buy box back first, then everything else

• Regained buy box control through inventory management and a rebuilt ad spend allocation, so the brand was winning the sale on its own listing before anything downstream was worth doing.
• Rebuilt the content — A+ content and video campaigns, produced with Beyond the Agency — so the traffic that arrived had a reason to convert.
• Drove advertising efficiency, cutting ACoS and lifting ROAS rather than chasing volume, with TACoS as the number that had to come down.
• Expanded the ad types into Sponsored Brands Video and Sponsored Display, with Brand Analytics informing where to push next.

The order matters here more than usual. Advertising into a listing you do not control means paying to send customers to a competitor sitting on your own detail page. Every dollar of that spend is worse than wasted, and it is the first thing to fix.

What changed

Revenue up 67% while ad spend fell by half

Monthly revenue moved from $15,000 across 2020 to more than $25,000 across 2021 — a 67% increase. Over the same period ad spend came down from more than $6,000 a month to under $3,000, roughly half of what the previous campaigns were consuming, and TACoS fell from over 35% to under 10%.

Those two directions together are the whole result. Revenue rising while ad spend falls means the growth is not being bought; a TACoS drop from 35% to under 10% is the same statement in the form the P&L cares about. Unit session percentage also rose 30% within the first three months, which is the earliest sign the content work was landing.

What we’d flag: these are annual averages compared against each other, which smooths over seasonality that a grocery brand genuinely has. The old report also mentions sales running at “almost 2x pace” in a two-month stretch; we have no baseline for that sentence, so it is not on this page.

Why it matters if you're reading this

You cannot advertise your way out of a lost buy box

Buy box loss is the quietest expensive problem on Amazon. Nothing breaks, no alert fires, and the brand keeps appearing in search — while a share of every sale you paid to generate goes to somebody who did not pay for it.

If your sessions look healthy and your conversion does not, check who is actually winning the sale on your listing before you touch the ad account. Fixing that first is what makes the advertising work worth doing at all.